Discussion:
0.7.4 is out.
(too old to reply)
Rob Landley
2017-06-19 18:29:53 UTC
Permalink
Raw Message
Up on the website and github, with release notes.

I spent a couple weekends fighting the help text parsing logic, but it
turns out the lifetime rules are all wrong for reusing hunks. I think I
need to rewrite about 1/3 of the help text parser to get it to do that.
So I just block copied the help text into the users in ps.c, which
should cut down on Elliott's bug reports. :)

Rob
enh
2017-06-19 20:30:42 UTC
Permalink
Raw Message
forgot to update the version number in main.c? (hmm... seems like the
tarball has that change, but git doesn't.)

should the CFG_TOYBOX_HELP bit in help_exit come _after_the verror_msg
call? it looks a bit odd saying...

See ps --help
ps: unknown option blah

...instead of

ps: unknown option blah
See ps --help

otherwise lgtm.
Post by Rob Landley
Up on the website and github, with release notes.
I spent a couple weekends fighting the help text parsing logic, but it
turns out the lifetime rules are all wrong for reusing hunks. I think I
need to rewrite about 1/3 of the help text parser to get it to do that.
So I just block copied the help text into the users in ps.c, which
should cut down on Elliott's bug reports. :)
Rob
_______________________________________________
Toybox mailing list
http://lists.landley.net/listinfo.cgi/toybox-landley.net
--
Elliott Hughes - http://who/enh - http://jessies.org/~enh/
Android native code/tools questions? Mail me/drop by/add me as a reviewer.
Rob Landley
2017-06-19 23:19:44 UTC
Permalink
Raw Message
Post by enh
forgot to update the version number in main.c? (hmm... seems like the
tarball has that change, but git doesn't.)
Sigh. Did you know git push --tags" doesn't update remote "master"? I
think this is at least the third time I've hit this...

(It pushed the commit so I could see it in the repo by hash, but master
didn't point to it. It was just sort of there.)
Post by enh
should the CFG_TOYBOX_HELP bit in help_exit come _after_the verror_msg
call? it looks a bit odd saying...
See ps --help
ps: unknown option blah
...instead of
ps: unknown option blah
See ps --help
Yeah, I second guessed this a couple times while implementing it.

Putting the error message on the last line is optimizing for veterans:
what's right above the command promopt is the cause of the failure. I
admit it's kind of awkward as english, but I'm assuming people only need
to be introduced to --help a couple times, and will be seeing error
messages over and over. So _usually_, the see --help isn't the
interesting bit. (It's vital for your first couple times, and beyond
that it's noise.) If the command itself output a lot of text, the error
message is harder to spot if it's not right at the end.

I also originally had it say "See %s --help for details" but that's more
english words that don't get translated, and the point is just to let
you know where to go next so you're not blocked. (I.E. --help is available.)

This is just my guess at an aesthetic issue, if you have a strong
opinion I'll change it. The other reading _is_ more natural as english,
and only very slightly less mechanically useful for programmers...

(It's always the questions with no clear winner that are hard to decide...)

Rob

Loading...